Recent advances in microwave imaging have enabled commercial development of microwave imaging systems that are capable of generating two-dimensional and even three-dimensional microwave images of objects and other items of interest (e.g., human subjects). At present, there are several microwave imaging techniques available. For example, one technique uses an array of microwave detectors (hereinafter referred to as “antenna elements”) to capture either passive microwave energy emitted by the target or reflected microwave energy reflected from the target in response to active microwave illumination of the target. A two-dimensional or three-dimensional image of a person or other item is constructed by scanning the array of antenna elements with respect to the target's position and/or adjusting the frequency (or wavelength) of the microwave energy being transmitted or detected.
Transmit and/or receive antenna arrays for use in transmitting and/or receiving microwave energy can be constructed using traditional analog phased arrays or binary reflector arrays, such as those described in U.S. pat. application Ser. No. 10/997,422, entitled “A Device for Reflecting Electromagnetic Radiation,” and Ser. No. 10/997,583, entitled “Broadband Binary Phased Antenna.” For either type of array, the largest addressable volume with the highest spatial resolution is obtained by choosing a small wavelength λ, densely filling the array with antenna elements such that the spacing between adjacent antenna elements in both directions is λ/2, and maximizing the two-dimensional area of the array. For example, if the array is a square of side L, an object located at a distance L from the array can be imaged with a resolution of approximately λ.
However, the number of antenna elements, and therefore the cost of the array, is proportional to (L/λ)2. This quadratic cost dependency is an obstacle to either scaling up the size of an array to increase the addressable field of view or reducing the wavelength to increase the resolution. As used herein, the term “addressable field of view” (AFOV) refers to the volume addressable with high resolution (i.e., the volume that can be resolved within some specified factor of the highest resolution).
One solution that has been suggested for the cost-resolution-AFOV problem is to use a sparse antenna array, instead of a dense antenna array. Since resolution increases with numerical aperture, which depends on the diameter and not the area of the array, an array with two or four antenna elements spaced L apart can achieve the desired resolution. However, sparse arrays produce multi-lobed antenna patterns. If the array is a traditional transmit phased array and 1≧s≧0 is the sparseness factor, Parseval's Theorem of Fourier analysis states that only s of the transmit power falls into an area that the originally dense (s=1) array of the same extent resolves. If the sparse array is a reflector array, and a transmit horn illuminates the full extent of the originally dense (s=1) array, the sparse array processes only s of the horn's power. Therefore, the efficiency factor (i.e., the transmitted fraction that fills the original area) is s2. If the reflector array is used to both direct microwave illumination towards the target and receive reflected microwave illumination from the target, the overall efficiency factor η=s4. For example, a 50% sparse reflector array produces a transmit-to-receive efficiency of 1/16=6.25%. Thus, as the sparseness of the array is increased, the signal loss increases as the fourth power.
The signal-to-noise (SNR) ratio of a sparse array also suffers the same s2 or s4 dependency. In addition, the background noise (often referred to as “clutter”) that results from stray radiation further decreases the SNR for sparse arrays for several reasons. First, the vacant area of the originally dense (s=1) array becomes a collective plane mirror that specularly bounces the radiation with a fill factor efficiency of 1−s. Second, the remaining (occupancy) area geometry generally produces sidelobes that change direction in a poorly controlled fashion as the antenna phasing changes. The sidelobe weight increases as the sparseness of the array increases. To the degree that these two factors increase system noise as the array becomes sparser, SNR will vary empirically as sa/(1−s)b, where a≈4 and b≈1. Thus, sparse arrays result in an increase in signal loss and a decrease in SNR.
Therefore, what is needed is a cost-effective antenna array design for use in microwave imaging that results in a high AFOV without high power-law increases in signal loss or decreases in SNR.